2GR-FE 07 Camry

[Rokassileika]

Yup, I absolutely see now, that this is a true art. It is just so weird. I first installed the injectors, calibrated everything (at least i thought so) the system was working very consistantly and engine power was stronger than ever. For a week. I press the throttle- it responds very well. Of course, the temperature was ~10-12°C outside. And then one day, the temperatures outside dropped to 4-5°C, it became very windy and I immediately felt how engine was not that responsive- actually it is dissapointing me again. For a week now as well. Given that LTFTs are that much negative it looks like that I should look into decreasing the multiplier, but then again. When it is warm outside, the engine gets less air than it does now so how does it make it run that happy? In theory it should run even worse? This is turning my head too much I'm starting to think that selling the car would make my life a little easier at the moment. But sometimes (on the warm days..) when I feel how much power this engine can produce i stop myself from selling it. The calculated engine load that i was checking was from OBD using the Autocom DS150E and Autocom software. I do have a Techstream but it shows the same. Or maybe my transmission is dying and on colder ambient temperatures doesn't work as expected (hydraulic pressure issues?), giving unnecesary load on the engine? But my MPG doesn't get worse, it is very stable. I'm just really lost at this point.

[LPGC]

Not sure what you mean about the weather changes, 4 to 5degC is warmer than -10 to -12C so going from -10 to +4 is an increase in temperature not a decrease?

You get different results from gas temp sensors depending on where in the rubber gas pipes they're fitted. If fitted close to the reducer they're quicker to respond and mroe accurately reflect the temp of gas coming out of the reducer, if fitted closer to injectors they're slower to respond and then come the questions - does this position more accurately tell the temp of gas reaching injectors or does it tell you more about the external temperature of the gas pipes? Some considerations are that gas flowing through rubber gas pipes will have its temperature affected by contact with the gas pipes it is flowing through, but gas flows through very quickly so how by how much does temp of rubber gas pipes actually affect temp of gas flowing through by the time it has reached injectors...

You control how much air the engine gets using the accelerator pedal. At idle the idle control valve or electronic throttle body automatically controls how much air the engine gets to maintain correct idle speed. The engine could make more power few with cold air but most of the time you're using only part throttle or the engine is only idling. Humidity can have a bit of an effect. But it is more likely the differences you're seeing are due to pressure, temperature and/or injector response differences.

[Rokassileika]

Sorry, it was +10 to +12 °C degrees, my bad. I'm still experiencing inconsistance in power output. Yesterday, my friend was driving behind me and he noticed, that once i pressed the accelerator to the metal a cloud of white smoke appeared. I should inform that there is a switch to gasoline set on 4k RPMs. Anyway this kind of fuming gives me an idea that the engine is working in too rich condition? Now the funny thing is that the weather hasn't changed at all it was same temperature yesterday and today, but today the engine was very responsive and quick. I believe it should mean that the engine wasn't working rich for today? What the heck am I fighting with then I would describe the engine working whatever mood it is in a day.

[LPGC]

It's difficult to do a remote diagnosis, we've never got all the necessary information. For example your friend said some white smoke came out when you put your foot down, we don't know if the engine and exhaust were still warming up or fully warmed up, weather temperature, humidity, etc. White smoke could be too much petrol, exhaust condensation turning to steam, oil burning, head gasket type problem. If it is petrol it could be because fuel trims at a high load area on LPG are lean so when you switch to petrol and put your foot down pinj pulses are affected and it overfuels but this would only happen if the petrol ECU does add fuel trim info for high load / open loop operation. We can't see pinj, ginj, fuel trims, temperature readings, etc. Engines do naturally make a bit more power when fed with cold air. Right at the beginning of this thread I said 2GRFE engines are prone to knock sensor problems, symptoms include fluctuating power under high load conditions.

[Rokassileika]

Can knock sensors cause the first lambda (O2) sensors to detect a rich mixture? For example, the MAF sensor is working properly, but the knock sensors are generating incorrect signals from the engine, leading to incomplete combustion? Also, could this be related to poor transmission performance, since as far as I know, timing retard is applied in the engine during gear shifts? And is it possible to observe any values in live diagnostics that would indicate these sensors are not working properly? Maybe the P0420 could be related to this?

[Rokassileika]

I'm starting to suspect my air intake box problems.. Thats what i got after 140 miles trip. + Very erratic gearbox operation.

[LPGC]

In theory if ignition timing is retarded the exhaust valve will open too early while the air and fuel are still burning, then whether or not the lambda sensor in the exhaust will detect the unburned air might depend on where in the exhaust (how far along the exhaust way from the exhaust valve) the lambda sensor is. Another name for a lambda sensor is oxygen sensor because that's what they do, they detect oxygen in the exhaust gasses, they don't detect unburned fuel or anything else just oxygen content. If the air and fuel are still burning it means there must still be some oxygen to burn and a lambda sensor close enough to the exhaust valve might detect that. If the lambda sensor were further away from the exhaust valve (further down the exhaust pipe) perhaps it would be more likely that the remaining air and fuel would burn before reaching the lambda sensor. A misfire can have the same effect, a misfiring cylinder means the air and fuel in that cylinder won't be ignited before being pumped into the exhaust but once it reaches the exhaust it will probably be ignited by flames coming from the other exhaust valves of cylinders that are not misfiring. Very often when an engine has a misfire the lambda sensor on the cylinder bank that has the misfire will give a lean mixture reading because it is reading the oxygen from the air in the cylinder that is misfiring, but not necessarily, and again it can depend on where in the exhaust the lambda sensor is (along with other conditions such as whether or not there's a flame front in the exhaust from other cylinders).

P0172 - system too rich bank 1
P0175 - too rich bank 2
P0101 - MAF sensor range/performance problem
P0420 - Catalyst efficiency below threshold bank 1

Ignore P0420 for now because you could expect to get P0420 when P0172 condition is present.

The other errors are not so clear cut. The MAF sensor error is for range OR performance. A range error implies a voltage signal that is out of range, so if the MAF were designed to output 1v for no airflow and 5V for 250hp's worth of airflow but the ECU found it reading 0v or 6v output that would be a range error. Performance error could be that the engine knows the engine is doing 2000rpm and the throttle is half open so the engine must be drawing (say) 40hp worth of air but the MAF signal says the engine is drawing 10hp or 60hp worth of air, the signal is still within range but the ECU knows it is wrong. If the MAF sensor was over-reading the amount of air flowing into the engine the ECU and the ECU believed the MAF sensor it would cause the engine to run too rich, which could cause the 0172 and 0175. If the ECU decided not to believe the MAF sensor it might switch to using a mode in which it ignores the MAF sensor and defaults to using internal tables that say at half throttle and 2000rpm inject this(x) amount of fuel because at least that will keep the engine running until the driver gets home or to a garage to have the MAF sensor fixed, but this table method won't be as accurate as the MAF method and could cause the P0172 and P1075 itself. The ECU might switch to the other mode if at some point it detects a range or performance signal, or it might not. But other things might trigger P0101 too, for example some ECU's might trigger P0101 when running in normal mode (using the MAF) even if range and performance seem OK but P0172 and P0175 have both been triggered... If both cylinder banks are running rich it does seem likely that the MAF is over-reading, maybe not to extent of unviable range/performance levels but because it implies that fuel trims on both cylinder banks have gone fully negative so it suggests the MAF is over-reading airflow by probably at least 20%.

If it runs OK on petrol without these codes the most likely reason for all of them is that the LPG system is calibrated too rich.

[Rokassileika]

Everything is very inconsistant in my situation. Sometimes the engine ran better on LPG than on petrol. And the fact that, let's say, one day i calibrate LPG multiplier map and the engine works really decent, the other day everything is off, switching from petrol to LPG starts giving different STFT values compared to the yesterday ones it somehow tells me that calibration might not be the culprit of this problem. Besides, engine still feels weak on petrol. When I bought the car that day it was flying like a rocket on petrol, after LPG installation I felt a really noticeable power loss compared vs when it didn't have LPG system. The installer told me that he broke a pcv hose that I later replaced with a different one. What else i noticed that there is 2 or 1 broken clips on air cleaner box (maybe the installer forgot to mention that he broke that too?). The filter does feel like it fits solid, but the clips that are missing might be letting some air through the gap (maybe causing over reading for maf). Another thing what I found out people have trouble on Toyotas and high negative LTFTs is called an air straightener. They fit aftermarket air intakes without air straightener material that knock off all MAF readings somehow. Or put an aftermarket air cleaner box and same happens. Other symptoms that I have- sometimes i can smell gasoline on cold starts, and when I drive the car more agressively it kind of revives. For some time. And of course, what got in my eye was significantly high calculated engine load that I talked about earlier. Maybe the MAF is dying and the first days I enjoyed with my new injector setup without saving fuel riding just finished the MAF off ..I guess I will have to check the simple things first for now (should have done this in the first place, I just always tend to think that it is always something much more complicated)

[Rokassileika]

Decided to just take a used MAF sensor from junk yard for very cheap. These were the instant results- +11 ltfts on both banks after 5 miles driving. Felt difference in gear shifts instantly. Engine load is now less in same conditions. Power is now consistant, but not sometghing wow.. Maybe ECU needs time to re-learn the trims.. My main question is why it was over-reading if it was so. Maybe ECU is bad burning the internals of the MAF eventually? I meam MAFs fail in other way- by under reading the air thats entering. Not sure for how long these numbers are going to stay. Kind of getting used to this- "that car is amazing today!" , "i might just wanna sell it right now"

[LPGC]

Good

[Rokassileika]

Some updates. Had some spare time and what I noticed that the curves on calibration map has changed. Did a re-calibration today, but I just cant set the very low load area correctly. It feels imposible since the STFTs switching from petrol to LPG are almost always different in one way or another. And sometimes it looks OK. Anyway looks like I could deal with middle load (5-7 ms) pretty well the STFTs don't fluctuate to very positive or negative sides. The high load looks okay too, but the curves are not joined. LPG curve is more to the left vs petrol, the STFTs tell me that this is the way they should be (maybe because of LPG switching to petrol at 4k RPMs function is on?). The LTFTs are always positive now +5..+11..+14, no more -25 LTFT haha . But what I noticed today was a very small whistling sound with the window down and only when the engine was not in operating temp.. Now my first thought is- a small vacuum leak somewhere? Maybe it is not letting me to fine tune the LPG?

[LPGC]

You know it makes sense to investigate the whistling sound, it could be a vacuum leak, a vacuum leak could be in a troublesome position such as on the threads of a 'spud' (brass LPG injection nozzle) in the manifold potentially making one cylinder run leaner than others. But if the whistling only occurs on LPG it could be the LPG pressure reducer vibrating at sound frequency, this would be a gas pressure whistle not a vacuum leak.

When you're calbrating you should be looking at STFT's and LTFT's. Sometimes at idle and low loads the difference between running on petrol and LPG (not least including the fact that LPG is injected as a vapour which takes up more volume in the inlet manifold than liquid petrol) can see the petrol ECU switch between LTFT ranges because there is a very slight difference in pumping losses and engine efficiency.. The petrol ECU controls the throttle (or IAV) to maintain the same idle speed regardless of which fuel you're running on but that throttle position can be effected by the differece in calorific value of the fuel (how much heat you get from burning a certain amount of fuel), the stochiometric ratio (how much air it takes to burn a certain amount of fuel) and the pumping loss (the engine acts like a pump, it has to suck air in and blow exhaust gas out). If it wasn't for pumping loss / efficiency differences we might expect / assume that running on LPG would need more intake air so the throttle would need to open a little more because LPG needs more air to burn the same weight of LPG as weight of petro.. but offsetting that there is the fact that LPG makes more heat by weight than petrol and less heat by volume (especially as LPG is in vapour form while petrol is in liquid form when injected). It isn't unusual to see idle MAF readings and MAP readings lower when running on LPG compared to petrol, and in a way this implies that the engine is idling more efficiently on LPG than on petrol. The lower MAF and/or MAP reading can be enough to see the petrol ECU apply a different LTFT range. To explain LTFT range - A simple hypothetical petrol ECU fuel map table might have 100 columns for RPM and 100 rows for MAF reading so 10000 cells in total but might only have 20 LTFT fuel trim ranges in total, so each LTFT would cover (apply to) on average 5000 cells of the fuel map. If you're just watching STFT's (and not watching LTFT's) you won't know if any slight difference in MAF/MAP readings between running on petrol and running on LPG sees the petrol ECU jump to a different LTFT range, a different LTFT range could have a different LTFT value stored in it. If for example one of the ranges had +10% LTFT stored in it and STFT was hanging around -10% and we add the LTFT and STFT together the total trim comes ot 0%. Or if we had -10% LTFT and +10% STFT the total trim is stlll 0%.Both of those situations would be OK and basically equal in terms of total fuel trim applied but if you were only looking at STFT's and not taking into account LTFT's you could make the mistake of assuming the mixture had changed by 20% when you switched between fuels when actually it was fine, you just hadn't taken into account the LTFT range shifting when you switched fuels. Of course you would still want to do further checks to make sure that fuelling on LPG was the same as fuelling on petrol, and you have to remember that the LTFTs have been learned so at some point (running on petrol or on LPG) it has been learned to be the value it is, you never want to see excessively low or high STFTS (or LTFTS) when running on either fuel or switching between fuels, but unless you watch for changes in LTFT range you could end up making the LPG map worse instead of better.

That was a simple example of fuel trim ranges and the petrol system map. In practice fuel trim ranges and the map are more complicated. There will probably be a different set of LTFT's applied to the map for different engine temperatures and intake air temperatures, etc. Some of the fuel trim ranges may only be applied for a short time during engine warmup. The petrol system goes closed loop at a relatively cool temperature when some 'choke effect' is still being applied (excess fuel factor to compensate for petrol that isn't burned when the engine is cool). If any petrol that is injected isn't getting burned when the engine is relatively cool, then you switch to LPG, you should expect the fuel trims to go negative after switching to LPG because the LPG will burn because (due to it being injected as a vapour and having a low boiling point) it will always mix perfectly with the air and burn. If fuel trims go negative when you've switched to LPG solely for the reason described here it is a fuel efficiency gain for LPG over petrol, the amount of negative fuel trim represents the excess fuel factor that was necessary when running on petrol with an engine that was still warming up. But you need to be sure what you're looking at, there could of course be other reasons why fuel trims go negative when you switch to LPG.

Looking at your screenshot above, LTFT's are at around +12% and STFT's are at around -12%. If that condition continued we'd expect that eventually LTFT's would be steered to be close to 0%, at which point STTS's would also likely stay close to 0%, so then we could add 0% LTFT to 0% STFT and the total trim would be 0%... As far as the petrol ECU is concerned that would be alnost perfect, the least likely situation to cause problems.. as long as when you switched fuels LTFT+STFT were still close to 0% and STFT didn't go too positive or too negative. But the idea when mapping an LPG system is to have fuel trims that are the same as would-be if you were running on petrol, at any temperature, at any rpm or engine load. It is possible to do this, perhaps with the exception of accounting for the petrol system haivng to apply some excess fuel factor during warmup, the best way to account for the petrol excess fuel factor is probably to use reducer temperature compensation but it's not always a good idea to use reducer temp compensation because there is a difference between an engine that has coolant temp at 30C on a cold day that has been run and is cooling down compared to an engine that has been started from cold on the same day and is warming up, the former engine (that has been run) will be heat soaked including the petrol in the fuel rail, so it' probably won't need any choke / excess fuel factor, the latter engine won't be heat soaked and probably will need some choke / excess fuel factor. If you try to compensate using the reducer temp compensation method it will still be correct in some situations and wrong in others.

Most engines apply most choke / eff effect (in percentage terms) around the ver low and low load areas of map, some VW's / Audis apply choke / eff effect more evenly across the whole map. On most vehicles (if you don't try to compensate for it / dial it out) you should only see trims go negative for a short time during warmup at around the low load area, not so much acros the rest of the nap. if you're seeing it happen across the whole map it is more lilely due to pressure / temperature / cold LPG injectors affecting fuelling than choke / eff effect but you have to remember that the trim ranges you see applied during warmup may not be the same trim ranges that are applied after warmup.

[Rokassileika]

Okay, I think that I was missing something while doing calibrations. I always thought that STFTs has to keep similar at different load moments- lets say I'm driving at middle loads on petrol, STFTs are +10 for ex., I switch to LPG and I need to see the STFTs +10 as well. But at the same RPM in that scenario there will be a different air volume (When switched on LPG) that will have a different LTFT stored at that moment and this is why I should look at LTFTs + STFTs to get near that 0 when adding both. Am I thinking correct? But my question is.. If I have LTFTs and STFTs in sight why would I need to switch between petrol and LPG to set the multiplier? I will just add up LTFTs and STFTs and look for it to be 0. I'm sorry if the question seems to be stupid, I'm in that phase now- the deeper I dig the less I understand. By the way, the reducer I have is working on ~1.2 bar at the moment and I have 45 HP per cylinder. Should I try 2.9 holes instead of 2.7 (according to the table)

[LPGC]

When the engine is fully warmed up and idling on LPG, what is Ginj (gas injector pulse duration) and Pinj (petrol injector pulse duration)?

A potential issue with fitting nozzles that are too big can be due to it meaning Ginj has to be shorter, the injectors are linear from around a certain Ginj and although good results can still be had by pulsing them for durations shorter than the linear range it can make calibration more difficult. I think I advised the relatively small nozzles for the engine power output to help make it easier for you to calibrate it around idle loads and also because I think we established that your pressure reducer probably isn't capable of supplying enough LPG (at high enough pressure) for the engine at full load and high rpm... At the other end of the scale (at the high load high rpm end) the nozzles have got to be big enough to be able to supply enough LPG for rich enough mixture for the engine within the 'window of opportunity' between 2 engine rotations (between a full 4 stroke cycle, the time between the inlet valve on one cylinder opening and opening again) but if your pressure reducer couldn't supply enough LPG for high load high rpm anyway we don't need to worry about the LPG injectors 'running out of window' because the pressure reducer can't flow enough gas for those conditions anyway.

There are some engines (e.g. that run Atkinson type cycles) for which it could be more of a disadvantage to have Ginj much higher than Pinj in certain conditions but that those aspects are not as relevant on your engine. Say again my thinking was to keep the calibration around idle relatively easy for you because with your reducer it won't be capable of running flat out on LPG anyway... But you could increase nozzle size to 3mm if Ginj seems high enough around idle (and high enough if you hold the engine at say 2500rpm and then slightly back off the throttle so the rpms fall slowly from 2500rpm to idle, this is the real condition when Ginj and Pinj will be shortest).

Incidentally, I never put any faith in the nozzle size charts. The nozzle size charts assume an engine cylinder that can make X bhp but assume they make that power at a certain RPM and assume a certain petrol injector flow rate. A bigger lower revving engine might make the same power per cylinder but at a lower RPM, the RPM effects the frequency of injection pulses which obviously affects how often the petrol and gas injectors have to turn off, the off time could be described as 'dead time' and since LPG injectors take longer to open and close than petrol injectors it means the higher the rpm a cylinder makes hp at the bigger flowing (or faster opening / closing, usually all 3) the injectors have to be. For the same reason (but to a lesser extent) a big low revving engine could make the same power with lower flowing petrol injectors than a small high revving engine too. A turbo / supercharged engine can make more power with (relatively) lower flowing injectors than would be needed for a normally aspirated engine that made the same hp . But all this said, if I were converting your engine from scratch I would fit 3mm nozzles and fit a reducer than was capable of flowing enough gas to provide the engine with rich enough mixture at full load high rpm - but it's no problem for me to calibrate the low load area. A lot of the 'iffy' and 'done in a day' installers fit / used to fit slow LPG injectors that had didn't become fairly linear until 4+ms of Ginj, so to make calibration easier for themselves they fitted relatively small nozzles, the small nozzles helped make calibration easy by pushing Ginj to 4+ or 5+ms at idle so they didn't need to worry or account for the injectors not being pulsed within their linear range at low loads, this made the mapping/calibraton easier for them but it also meant the LPG install couldn't supply enough LPG for rich enough mixture at high rpm high loads, they got around that by simply setting their installs to switch back to petrol at high rpm or other similar get-arounds so they could get away with less work or use poor components, usually both.

[Rokassileika]

Going to keep writing here for now.. But theres have to be some sort of explanation. I do see that LPG temperature has difference in 9 degrees of celsius. But oh come on.. On petrol on idle stft + ltft is arround 10. I set multiplier, I got it correct, on LPG I get same readings (arround +10 combined). And then... BOOM.. LPG has got it at 4-5, I keep it for some time and it goes to even zero. Switch to petrol- gets back to +10 eventually. And that small shake (misfire) from petrol to LPG happens, but if I get it right for a very very precise and exact moment- there is no small misfires. Really getting out of breath. What kind of nonsense is this?

There was a switch-over time set to 200, I set it to 150, didn't help. AND I noticed that my RPM correction screen just vanished!!

I did eliminate the active mount vacuum lines, closed it well, there was a change in fuel trims and no difference how that mount behaves.

I do have a leaking bank 1 valve cover. Can it really make that situation that bad? Why I still haven't given up

[Rokassileika]

Update:

Found a broken clip on air filter box where part of the box sits on the filter sealing the gap. Fixed it with my DIY clip. I can clearly see now true corrections at idle, no outrageous fluctuations. It is clear now what has to be adjusted. I believe it could have been a turbulence issue? Even though the gap looked solid and the clip looked like it would make no change it seems that it still did.

Anyway..

On petrol combined fuel trims are still +10.. +12. When idling I would consider it is still high

Bought a smoke machine to rule out any other vacuum leaks if exists, waiting for delivery.

Until that time I still have lots of issues with EW10J4. The trims on petrol are even higher, could even be a weak fuel pump.

[Rokassileika]

Update.

Found no obvious vacuum leaks, everything seemed to look solid enough, maybe very minor leak under higher pressure from smoke tester..

And I always thought about this- when I bought the car it was very powerful, constant. All this started after the installation of LPG.

This is what I found, please note that there is NO high pressure, airways are open.

https://youtube.com/shorts/BfRJC-p8Ueg? ... ORuPEwSZcM

The installer probably broke these clips and didn't pay attention. I did notice that before, but never thought it could be that bad..

[LPGC]

The leak is from the air filter housing. Since any MAF would be downstream towards the engine from the air filter housing the leak in your video wouldn't affect any MAF readings and isn't a vacuum leak, it is just a leak on the air filter housing which at worst might mean a small minority of air the engine uses doesn't go through the air filter. It will be inconsequential.

[Rokassileika]

I've heard people having problems with cone air filters without honey combs on these engines- the turbulence arround the MAF sensor creates a rich condition. I thought maybe this could be a possibility, just the opposite. Anyway, I was randomly scrolling facebook feed and noticed an identical situation- inconsistant power output for a 2009 Camry with the 2.5 engine and LPG. Except, he drove the car to like 15 mechanics and LPG specialists and was told to look for the problem elsewhere.. To be sepcific- nobody can find the problem. So do I. Theres really gotta be something specific with this car. The knock sensors? But what about the 2.5? Are they known for the same issue? And why this happened right after the installation of LPG? This only tells me that there is an error of the installer that left me with for all these years, yet I can't find where it is.

But to note:

I did put a new Zavoli Zeta S reducer- the p0420 was gone for a few weeks then came back

New Barracuda injectors did something- the car was feeling more alive for one week only

Driving the car agressively from the very beginning (cold engine) fixes the issue, but if i start driving it slowly it does all that stupid thing again

Fuel trims are erratic at idle only, sometimes stable. And kind of really high on petrol i would say
..
Maybe clogged cats? Lambdas dying? Leaking petrol injectors? Any ideas?

Should i try some cheap knock sensors to see if it does anything?

Any advice is golden for me as I say. What would You do in my situation (except selling the car hehe )